hail hazard and risk assessment in europe and the relation ...tippett/scc2/presentations/day...
TRANSCRIPT
KIT – The Research University in the Helmholtz Association www.kit.edu
Institute of Meteorology and Climate Research – Troposphere Research (IMK-TRO)
Hail hazard and risk assessment in Europe and the relation to orographic & atmospheric characteristics Michael Kunz, Heinz Jürgen Punge, Elody Fluck, Manuel Schmidberger, Susanna Mohr, David Piper, and Marc Puskeiler
2016 Severe Convection and Climate Workshop
Severe hailstorms 2013: Andreas Radar composite (radars IMK + DWD, cell tracking TRACE3D) 28 July 2013
triggering
supercell formation by cold pool of single cell
Lon (°)
Lat (
°)
heig
ht
Hail hazard and risk assessment [email protected] 9th March 2016
2016 Severe Convection and Climate Workshop
Severe hailstorms 2013 Germany
2-4 cm 6-7 cm > 8 cm
27 July 28 July
Data:
6 August
Hannover Wolfsburg
Stuttgart
Hail hazard and risk assessment [email protected] 9th March 2016
© Marco Kaschuba
14 cm
Total loss: $ 4.6 bn insured loss 27/28 July: $ 3.1 bn
(Swiss Re, 2014)
2016 Severe Convection and Climate Workshop
Questions…
Regional-scale hail frequency?
Reasons for the spatial distribution?
Long-term variability of hail potential?
Hail hazard and risk assessment [email protected] 9th March 2016
2016 Severe Convection and Climate Workshop
Hail growth within convective clouds
Condensation process by cloud condensation nuclei CCN
Nucleation by IN: ice crystal
hailstone formation by fast riming
Riming by accretion of supercooled droplets: hail embryos (graupel, frozen drop)
(Straka, 2009; Pruppacher and Klett, 2010; Houze, 2014)
Hail hazard and risk assessment [email protected] 9th March 2016
condensation
collision- coalescence Raindrops
nucleation/ freezing
accretion
rapid accretion
-37°C
cloud- base
embryo formation
hail formation
0°C
2016 Severe Convection and Climate Workshop
-37°C
cloud- base
Hail growth within convective clouds
embryo formation
hail formation
(Straka, 2009; Pruppacher und Klett, 2010; Houze, 2014)
Hail hazard and risk assessment [email protected] 9th March 2016
0°C
Hailstreaks: small extent (∅: 40 km²; Changnon, 1970)
⇒ surface stations not sufficient for monitoring Proxies for hail detection: - radar reflectivity - overshooting top
2016 Severe Convection and Climate Workshop
Hail proxy: Overshooting tops OTs: intrusions of convective cloud to lower stratosphere Indicator of very strong convective updrafts
Hail hazard and risk assessment [email protected] 9th March 2016
Combination Satellite (MSG SEVIRI) and Reanalysis (ERA-Interim)
2004 – 2014
Events per year and 100 km²
2016 Severe Convection and Climate Workshop
v
v
Hail hazard and risk assessment [email protected] 9th March 2016
Hail proxy: Radar reflectivity Radar: high spatial / temporal resolution, available for several European countries Dual-pol only over recent years
number density
Methods:
Maximum reflectivity
Vertical extent (Waldvogel et al., 1978)
Cell tracking (TRACE3D), advection correction
Clutter correction (lightning), calibration (insurance)
2016 Severe Convection and Climate Workshop
Hail detection radar vs damage data Calibration: categorical verification with skill scores Building insurance data;summer half-year 2005 - 2011
2D: threshold exceedance
Ski
ll sc
ore
Ski
ll S
core
Max{HSS} = 0.61 Max{HSS} = 0.71
3D: vertical extent
9th March 2016 Hail hazard and risk assessment [email protected]
Reflectivity in dBZ HK in km
(Puskeiler and Kunz, 2016)
2016 Severe Convection and Climate Workshop
Hail frequency in European countries
Number of days with hail signals (2D/3D radar refl.)
Belgium, Luxembourg 2004-2013
" 1 - 5" 6 - 8" 9 - 11" 12 - 14" 15 - 17" 18 - 21" 22 - 29
(Fluck et al., 2015; Kunz and Schmidberger, 2014)
Hail hazard and risk assessment [email protected] 9th March 2016
France 2005-2013
Germany 2005-2011
2016 Severe Convection and Climate Workshop
Hail probability related to orography Flow dynamics: Flow around regime, formation of low-level flow convergence
Froude number Fr
rela
tive
frequ
ency
in %
HNUFr =
(Kunz and Puskeiler, MZ, 2010)
Hail hazard and risk assessment [email protected] 9th March 2016
Histogram of Fr on hail days SW Ger.
2016 Severe Convection and Climate Workshop
Hail probability vs orography Semi-idealized COSMO-DE simulations (2.8 km) Initialization with ambient conditions that favor hailstorms (Fr ~ 0.6)
Flow convergence Convective available potential energy CAPE
CON
9th March 2016 Hail hazard and risk assessment [email protected]
(Köbele, 2014)
DIV
2016 Severe Convection and Climate Workshop
Hail hazard and risk assessment [email protected] 9th March 2016
Hail probability vs orography COSMO sensitivity studies using different model setups
Ref
(Brombach, 2012)
2016 Severe Convection and Climate Workshop
Hail hazard and risk assessment [email protected] 9th March 2016
Risk Assessment: CAT-Modeling Hail risk for the Insurance Industry
Risk = hazard x exposure x vulnerability
(IPCC SREX, 2012)
Event catalogue
Stoch. Modeling e.g. 10 000 years
intensity
degr
ee o
f dam
age
Portfolio
10%
Damage function
Portfolio, object data
(Daniell, 2013)
2016 Severe Convection and Climate Workshop
Long-term changes in hail probability
Warming due to climate change
Increased evaporation
Invigoration lifting / convective energy
Changes frequency / intensity events
Dynamics, weather patterns
?
?
Hail hazard and risk assessment [email protected] 9th March 2016
2016 Severe Convection and Climate Workshop
Trends Convective Parameters Trends 90% percentiles from radio-soundings Increase in convective energy caused by increase in moisture (low lev)
Soundings: 1978-2009
(Mohr and Kunz, ARE, 2013)
Hail hazard and risk assessment [email protected] 9th March 2016
Percentiles estimated from pdf (Gamma / Weibull distrib.) Eliminate of autocorrelation: trend-free pre-whitening
2016 Severe Convection and Climate Workshop
Logistic regression Logistic Hail Model LHM
Potential Hail Index (PHI)
mit
Logistic Hail Modelling
Mohr, Kunz, and Keuler, JGR, 2015) (Mohr, Kunz, and Geyer, GRL, 2015
SLI: Lifted Index Tmin: Min. Temp. 2 m T2m: 12 UTC Temp. 2 m oWL: objective weather patterns
9th March 2016 Hail hazard and risk assessment [email protected]
2016 Severe Convection and Climate Workshop
Hail hazard and risk assessment [email protected] 9th March 2016
Hail potential in Europe (1951-2010) Combination of various meteorological parameters relevant for thunderstorm / hail: Potential Hail Index PHI
1951-2000
Reanalysis, driven by NCEP–NCAR 1 (@ HZG)
II decr
ease
in
crea
se
(Mohr, Kunz and Geyer, GRL, 2015)
not significant
trend mean
2016 Severe Convection and Climate Workshop
1951-2000
Reanalysis, driven by NCEP–NCAR 1 (@ HZG)
Hail hazard and risk assessment [email protected] 9th March 2016
Hail potential in Europe (1951-2010) Combination of various meteorological parameters relevant for thunderstorm / hail
II
(Mohr, Kunz and Geyer, GRL, 2015)
mean
2016 Severe Convection and Climate Workshop
PHI in the future (Germany) de
crea
se
inc
reas
e
Mini-Ensemble based on 7 RCMs different SRES, realizations, RCMs
2021-2050 vs 1971-2000
(Mohr, Kunz and Keuler, JGR, 2015)
Hail hazard and risk assessment [email protected] 9th March 2016
2016 Severe Convection and Climate Workshop
Conclusions Large damage potential associated with hail in Europe
Hail hazard assessment from remote sensing instruments:
− robust and physically plausible
− OT approach: consistent for larger areas
− Radar-derived signals: detailed estimates
Hail signals: high spatial variability related to
− overall climatology (→ stability)
− (low) mountains (→ preferred location downstream)
Hail potential has been increased over past decades, (slight) increase in the future; trends statistically not significant
Hail hazard and risk assessment [email protected] 9th March 2016
2016 Severe Convection and Climate Workshop
…Thank you Hail hazard and risk assessment [email protected] 9th March 2016
19-21